Energy Expenditure

Energy expenditure is the total caloric cost of running a human body over a period of time. Most commonly reported over 24 hours as total daily energy expenditure (TDEE), but also meaningful at shorter timescales — resting energy expenditure over an hour, exercise energy expenditure during a training session, or thermic effect of a single meal.

The partitioning

For a sedentary adult, a typical breakdown:

• BMR (basal metabolic rate): 60–70% of TDEE
• TEF (thermic effect of food): 8–12%
• NEAT (non-exercise activity thermogenesis): 15–25%
• EAT (exercise activity thermogenesis): 0–10% depending on training

For a highly active athlete, the partition shifts substantially toward the activity components — an endurance athlete training 15 hours/week may have 25–40% of TDEE coming from EAT, with BMR correspondingly less dominant.

How each component is measured or estimated

• BMR / RMR: indirect calorimetry in the lab; predictive equations (Mifflin-St Jeor, Katch-McArdle) in consumer apps.
• TEF: measured by post-prandial calorimetry; estimated at 8–12% in most calculations, higher for high-protein diets.
• NEAT: estimated from accelerometer-based step counts, movement intensity, and (in wearables) heart-rate data. Tends to undercount low-level fidget movement.
• Exercise EE: heart-rate-based estimates from wearables, MET-based estimates from activity logs, direct measurement via portable calorimetry.

What consumer wearables report — and don't

Wearables like Apple Watch, Fitbit, Garmin, Whoop, and Oura report "calories burned" as a daily total, typically summing an estimated BMR with movement-derived activity calories. Published validation work consistently finds:

• Step counts are reasonably accurate (±3–10% depending on device and conditions).
• Active-calorie estimates carry ±10–25% error at the individual level.
• Total daily energy expenditure estimates tend to run 10–20% lower than doubly-labelled-water reference values in many studies.

This is directional information, not a measurement. Treating a wearable's "calories burned" number as ground truth for caloric-balance planning routinely produces misfit: users end up in larger or smaller deficits than they believe.

Adaptive shifts

Sustained caloric deficit reduces total energy expenditure below predicted via:

• BMR suppression (adaptive thermogenesis).
• NEAT drop (the largest component of the adaptive response).
• TEF reduction as digestive efficiency improves.
• Reduced spontaneous activity choices.

Aggregate effect: 200–500 kcal/day below predicted after 8–12 weeks in a meaningful deficit. A tracking app that does not adapt its TDEE estimate to the user's actual weight-trend data will increasingly diverge from reality as the cut progresses.

Tracking implication

The "Calories Out" side of CICO is always an estimate. The most actionable approach: use wearable and app estimates as starting points, calibrate against weekly-average-weight-trend data over 2–4 weeks, and recalibrate every 4–6 weeks during active body-composition change. No single number is ground truth; the empirical fit to the actual trend is.